基于傅立叶变换的针织物密度分析

Analyzing of Knitting Fabric's Density Based on DFT

对如何使用离散傅立叶变换方法分析针织物密度进行了探讨.由于经过后整理的针织物线圈不均匀,各线圈纵行或各线圈横列只是近似地处在纵行线或横列线上,而且纵行线与横列线也不垂直,用二维离散傅立叶变换方法处理针织物灰度图像f(x,y)得到的结果误差较大.但采用对f(x,y)逐列相加后对其和进行一维傅立叶变换及对f(x,y)逐行相加后对其和进行一维傅立叶变换的方法能得到较准确的结果.

When one analyzes knitting fabric construction's digital image, the knitting density is one of the key parameters. This paper explores how to use DFT (The Dispersed Fourier Transform) to analyze knitting fabric's density. After backprocessing the coils of the knitting fabric are not always welldistributed, the rows/columns are not just on the rows'/columns' lines, and the rows' and columns' lines are not vertical. Therefore, if 2DDFT (The 2 Dimensionality Dispersed Fourier Transform) is used to analyze the knitting fabric's grayscale image--f(x,y), the error of the result is always too large. However if we first add the f(x,y)'s every row or add the f(x,y)'s every column, and then use DFT to transform the adds, the right result can be obtained.